📰 DiffuSpace, founded by the HKU Dream team, open-sources Clover-35B-A3B, a diffusion LLM hitting 680 tokens/s on one GPU and 73.6 on SWE-bench Verified
On Oct 28 DiffuSpace, whose members come from the HKU NLP group behind Dream 7B, released its first model, Clover-35B-A3B-Preview: a mixture-of-experts diffusion language model initialized from Qwen3.6-35B-A3B, with 35B total and about 3B active parameters, a 128K context window and block-parallel…
💬 What's your take? Join the discussion in the comments.
Read the original Chinese version
📰 港大 Dream 团队创立的 DiffuSpace 开源 Clover-35B-A3B 扩散语言模型：单卡 680 tokens/s，SWE-bench Verified 73.6
扩散智能 DiffuSpace（成员来自推出 Dream 7B 的港大 NLP 实验室）10 月 28 日发布首个模型 Clover-35B-A3B-Preview：基于 Qwen3.6-35B-A3B 初始化的 MoE 扩散语言模型，总参 35B、激活约 3B、128K 上下文，按块并行去噪生成。
💬 你怎么看？欢迎在评论区讨论。